We present new sets of pion and kaon fragmentation functions obtained in NLO combined analyses of single-inclusive hadron production in electron-positron annihilation, proton-proton collisions, and deep-inelastic lepton-proton scattering with either pions or kaons identified in the final state. At variance with all previous fits, the present analyses take into account data where hadrons of different electrical charge are identified, which allow to discriminate quark from anti-quark fragmentation functions without the need of non trivial flavor symmetry assumptions. The resulting sets are in good agreement with all data analyzed, which cover a much wider kinematical range than in previous fits. An extensive use of the Lagrange multiplier technique is made in order to assess the uncertainties in the extraction of the fragmentation functions and the synergy from the complementary data sets in our global analysis.

Abstract. This White Paper presents the science case of an Electron-Ion Collider (EIC), focused on the structure and interactions of gluon-dominated matter, with the intent to articulate it to the broader nuclear science community. It was commissioned by the managements of Brookhaven National Laboratory (BNL) and Thomas Jefferson National Accelerator Facility (JLab) with the objective of presenting a summary of scientific opportunities and goals of the EIC as a follow-up to the 2007 NSAC Long Range plan. This document is a culmination of a community-wide effort in nuclear science following a series of workshops on EIC physics over the past decades and, in particular, the focused ten-week program on "Gluons and quark sea at high energies" at the Institute for Nuclear Theory in Fall 2010. It contains a brief description of a few golden physics measurements along with accelerator and detector concepts required to achieve them. It has been benefited profoundly from inputs by the users' communities of BNL and JLab. This White Paper offers the promise to propel the QCD science program in the US, established with the CEBAF accelerator at JLab and the RHIC collider at BNL, to the next QCD frontier. Preamble Editors' note for the second editionThe first edition of this White Paper was released in 2012. In the current (second) edition, the science case for the EIC is further sharpened in view of the recent data from BNL, CERN and JLab experiments and the lessons learnt from them. Additional improvements were made by taking into account suggestions from the larger nuclear physics community including those made at the EIC Users Group meeting at Stony Brook University in July 2014, and the QCD Town Meeting at Temple University in September 2014.Abhay Deshpande, Zein-Eddine Meziani and Jian-Wei Qiu November 2014 Editors' note for the third edition Since the 2nd release of this White Paper, the NSAC's Long Range Plan (2015) was successfully completed. The EIC is a major recommendation of the US nuclear science community. In the current release (version 3) we have fixed some minor remaining errors in the text, and have added a few new references. While the core science case for the EIC remains the same, the machine designs of both options, the eRHIC at BNL and the JLEIC at JLab keep evolving. In this 3rd release of the EIC White Paper instead of making substantial changes to the machine design sections (5.1 and 5.2), we give references to the most recent machine design documents.

We discuss techniques and results for the extraction of the nucleon's spin-dependent parton distributions and their uncertainties from data for polarized deep-inelastic lepton-nucleon and proton-proton scattering by means of a global QCD analysis. Computational methods are described that significantly increase the speed of the required calculations to a level that allows one to perform the full analysis consistently at next-to-leading order accuracy. We examine how the various data sets help to constrain different aspects of the quark, antiquark, and gluon helicity distributions. Uncertainty estimates are performed using both the Lagrange multiplier and the Hessian approaches. We use the extracted parton distribution functions and their estimated uncertainties to predict spin asymmetries for high-transverse momentum pion and jet production in polarized proton-proton collisions at 500 GeV center-of-mass system energy at the Relativistic Heavy Ion Collider at Brookhaven National Laboratory, as well as for W boson production.

Polarized deep inelastic scattering ͑DIS͒ data are analyzed in leading and next-to-leading order of QCD within the common ''standard'' scenario of polarized parton distributions with a flavor-symmetric light sea ͑antiquark͒ distribution ␦q , and a completely SU͑3͒ f broken ''valence'' scenario with totally flavorasymmetric light sea densities (␦ū ␦d ␦s). The latter flavor-broken light sea distributions are modeled with the help of a Pauli-blocking ansatz at the low radiative or dynamical input scales of LO͑NLO͒ 2 ϭ0.26 (0.40) GeV 2 which complies with predictions of the chiral quark-soliton model and expectations based on the statistical parton model as well as with the corresponding, well established, flavor-broken unpolarized sea (d Ͼū ). Present semi-inclusive DIS data cannot yet uniquely discriminate between those two flavorsymmetric and flavor-broken polarized light sea scenarios.

A next-to-leading order QCD analysis of spin asymmetries and structure functions in polarized deep inelastic lepton nucleon scattering is presented within the framework of the radiative parton model. A consistent NLO formulation of the Q 2 -evolution of polarized parton distributions yields two sets of plausible NLO spin dependent parton distributions in the conventional MS factorization scheme. They respect the fundamental positivity constraints down to the low resolution scale Q 2 = µ 2 N LO = 0.34 GeV 2 . The Q 2 -dependence of the spin asymmetries A p,n,d 1 (x, Q 2 ) is similar to the leading-order (LO) one in the range 1 ≤ Q 2 ≤ 20 GeV 2 and is shown to be non-negligible for x-values relevant for the analysis of the present data and possibly forthcoming data at HERA.